Removing Rows that Are Exclusively NA

Question

Currently I have data that looks like this (just with significantly more NA filled columns and significantly more rows that are exclusively NAs):

Column 1	Column 2	Column 3
NA	NA	NA
Texas	Oklahoma	NA
NA	Florida	Florida
NA	NA	NA

I'd like for it to look like this

Column 1	Column 2	Column 3
Texas	Oklahoma	NA
NA	Florida	Florida

I don't want to get rid of all rows that have an NA value, I just want to get rid of all rows that have nothing but NA values.

Thanks in advance.

data[complete.cases(data), ] – MatthewR Oct 17 '21 at 23:54 — MatthewR, Oct 17 '21 at 23:54

score 1 · Answer 1 · answered Oct 18 '21 at 01:12

1

You can use janitor package:

janitor::remove_empty(dat, which = "rows")

answered Oct 18 '21 at 01:12

Bloxx

1,495
1
9
21

score 0 · Answer 2 · answered Oct 18 '21 at 00:06

0

If you want to get rid of rows that contain only NA but keep rows that have some NA values you could combine is.na and rowSums

dat <- dat[rowSums(!is.na(dat)) > 0, ]

is.na will return a TRUE or FALSE which will be treated as 1 and 0 for the rowSums. Any row with a value that is not NA will have a sum greater than 0.

answered Oct 18 '21 at 00:06

Tjn25

685
5
18

2

or you can use janitor package `janitor::remove_empty(dat, which = "rows")` – Bloxx Oct 18 '21 at 00:08
@Bloxx that uses `is.na` and `rowSums` just inside another package you have to install. – Tjn25 Oct 18 '21 at 00:14
you are right. however, janitor package is very useful with data cleaning. Also with cleaning column names. – Bloxx Oct 18 '21 at 00:22
1

@Bloxx, you could post that as an answer ... – Ben Bolker Oct 18 '21 at 01:07

Removing Rows that Are Exclusively NA

2 Answers2